Integrative Multiomics Mendelian Randomization Prioritizes Endoplasmic Reticulum Stress‐Related Genes Associated With Gastric Cancer Risk
Dong Yang, Yuanlin Liu, Xiangyu Meng, Chao Wang, Zhiyong Yu, Tao ZhangBackground
Emerging research indicates connections between endoplasmic reticulum (ER) stress and gastric cancer (GC), but related genetic factors remain unclear. This study is aimed at prioritizing ER stress‐related genes associated with GC risk using multiomics data.
Methods
We integrated GC GWAS data from FinnGen (discovery) and United Kingdom Biobank (replication) cohorts with QTL datasets, including mQTLs, eQTLs, and pQTLs. Summary‐data‐based Mendelian randomization (SMR) and colocalization analyses were performed to prioritize candidate genes, followed by characterization using transcriptomic profiling (GSE118916), methylation‐expression correlation analysis (TCGA‐STAD), and single‐cell RNA‐seq analysis (GSE270680).
Results
SMR analyses identified 83 mQTL, 10 eQTL, and 2 pQTL signals associated with GC risk. Among these, colocalization evidence was observed for 47 mQTL, 5 eQTL, and 2 pQTL signals. Transcriptomic analysis showed that PCSK6 , YOD1 , and RNF5 exhibited expression changes directionally consistent with their SMR associations with GC risk, whereas CTNNBIP1 showed an inconsistent expression pattern. PCSK6 was supported by mQTL, eQTL, and mQTL‐eQTL evidence but lacked colocalization. YOD1 , identified through pQTL‐SMR, showed strong colocalization evidence (PPH4 = 0.95). RNF5 was highlighted via eQTL‐SMR but lacked colocalization support. CTNNBIP1 was supported by mQTL, eQTL, and mQTL‐eQTL evidence and colocalization, with a significant methylation‐expression correlation observed in TCGA‐STAD. Single‐cell RNA‐seq revealed that PCSK6 was predominantly expressed in fibroblasts, YOD1 in epithelial cells and mast cells, and RNF5 broadly across fibroblasts, epithelial cells, and selected immune cells.
Conclusions
This multiomics study prioritized PCSK6 , YOD1 , and RNF5 as ER stress‐related candidate genes showing genetic associations with GC risk and directionally consistent expression patterns, with single‐cell analysis indicating distinct cell‐type‐specific expression patterns. These findings offer useful context for future functional and mechanistic investigations of the prioritized candidate genes.